Srujana Rangapuram

• Designed a model interface for better LLM inference and RAG support using Ollama, Langchain, FastAPI, and Docker. • Improved LLM response accuracy and hallucination by 15% with prompt engineering, guardrails, and data embedding. • Enhanced big data ETL pipelines and 3D visualizations for biological data from Google Data Commons and USDA API. • Developed a Streamlit feed player to predict possibilities across categories like Jobs, Bee population, and wages.

• Automated Computer Vision projects with AI ML tools (Transformers, PyTorch, TensorFlow), saving 950+ manual hours. • Developed custom Generative AI models (GPT, Llama, Gemma), boosting business performance by 50,000 USD through enhanced data cleaning, data embedding, text generation, and summarization. • Led statistical modeling projects, like time-series forecasting and survival analysis, boosting client operations by 50%. • Optimized anomaly detection and image segmentation models with ResNet and YOLO, achieving 80-93% precision. • Increased AI model accuracy by 10-35% through efficient data analysis, feature selection, fine-tuning, and testing. • Revamped UX/UI and web applications utilizing React, Microsoft Azure, JavaScript, PostgreSQL, Git, and Figma.

• Devised a Python NLP chatbot with Rasa for rapid IT ticket creation and query resolution, saving 5 minutes per user. • Extracted insights for business KPIs and project solutions through big data ETL pipelines and integrated them with existing applications. • Analyzed healthcare datasets (Insurance Claims, EHR, Clinical Trials) using tools like Snowflake, Power BI, Hadoop, Microsoft Excel, and PostgreSQL. • Completed a highly immersive Data Science program covering data manipulation, model development with ML algorithms, and visualization with tools like Python, SQL, NoSQL, MongoDB, and Hadoop.

• Designed healthcare and customer service applications using Angular, TypeScript, Java SpringBoot, Git, REST APIs, Kafka, Grafana, and Jenkins. • Led the CI/CD automation by cross-functional collaboration to troubleshoot challenges, boosting efficiency by 60%. • Design and implementing testing scripts for unit testing, Integration testing, End-to-end testing and A/B testing. • Improved code modularity and technical documentation with ad-hoc libraries, better coding standards and regular code reviews.